Principle Component Analysis of groundwater level-an example for Choshui River alluvial fan
| Autor: | Sheng-ChiehChen, 陳聖傑 |
|---|---|
| Rok vydání: | 2012 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 100 This research used PCA (Principle Component Analysis) to analyze month average groundwater level from 2007 to 2010 at Choshui River alluvial fan. The four principle components were extracted from the groundwater level of each aquifer of Choshui River alluvial fan. These components were used to compare to the local hydrology phenomenon and to conjecture the driving force of groundwater level change. The PCA was also applied to analyze the short-term groundwater level change during each typhoon at Choshui River alluvial fan. Thirteen rainfall intense from 2007 to 2009 was selected. The first principle component was extracted from the groundwater level of each aquifer of Choshui River alluvial fan in order to understand groundwater recharge due to rainfall. Results showed that in first aquifer, the first and second principle components can be represented as the rainfall recharge to aquifer, and third and fourth principle component can be represented as the effect of downstream land subsidence to the groundwater level. In second aquifer, the first and third principle components can be represented as the rainfall recharge to aquifer, and second and forth principle component can be represented as the effect of downstream land subsidence to the groundwater level. In third aquifer and fourth aquifer, also the first and third principle components can be represented as the rainfall recharge to aquifer, and second and forth principle component can be represented as the effect of downstream land subsidence to the groundwater level. The four principle components in every aquifer can be used to explain more than 85 % of influence ratio of the original data. This show that the change of groundwater level in Choshui River alluvial fan is most significant in terms of rainfall recharge to aquifer, and is then in terms of effect of downstream land subsidence. Moreover, results showed that the first principle component can explain the original data more than 70% to 80% when first principle components were extracted from every aquifer of thirteen typhoons from 2007 to 2009. It makes an implication that except of first principle component, the remain principle components explain less original data, and the trends of remain components become signals hard to identify. The result also indicated that the scores of first principle component in every aquifer had a trend to lift after several hours and then maintained a fixed uplift trend when the typhoon was occurred. This implied that the real effect of rainfall recharge caused groundwater level change maintain a fixed velocity. This research confirmed that PCA analysis using in short-term typhoon episode can filter weight load caused by the rainfall and can find the actual percentage caused by rainfall recharge groundwater level change. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|